Evaluation and Comparison of 11 Sets of Gridded Precipitation Products over the Qinghai-Tibet Plateau

Pinzeng Rao,Fang Wang,Xing Yuan,Yang Liu,Yang Jiao
DOI: https://doi.org/10.1016/j.atmosres.2024.107315
IF: 5.965
2024-01-01
Atmospheric Research
Abstract:The Qinghai Tibet Plateau (QTP) poses challenges in developing high -quality gridded precipitation data with high spatial and temporal resolution due to its complex terrain, harsh weather conditions, and sparse gauge observations. Previous research showed that gridded precipitation products from remote sensing satellites, reanalysis, and merged data are useful over the QTP. However, the precipitation products were usually calibrated against ground meteorological stations, so the evaluation based on these stations might be misleading. This study collected 11 sets of gridded precipitation products on the QTP, and evaluated them by using 53 independent rain gauges obtained from the hydrological yearbook (HY). The results showed that the CMFD precipitation product performed better than other products at meteorological sites from the National Meteorological Information Center (NMIC), with average daily and monthly correlation coefficients (CCs) of 0.55 and 0.94 and root mean square errors (RMSEs) of 3.78 and 0.44 mm/d, respectively. Following closely is CLDAS. In contrast, MSP performed best at HY sites, with average CCs of 0.70 and 0.91 and RMSEs of 2.47 and 0.65 mm/d, respectively. The following are TPHiPr, GSMAP, and CLDAS. Some global products, such as CHIPRS and CMORPH, performed poorly over the QTP. Most products overestimated the frequency of light precipitation (0.1- 0.5 mm/d) and underestimated the frequency of heavy precipitation. The fluctuation amplitudes of relative deviations for all precipitation products are markedly greater at hydrological sites compared to that at meteorological sites. Therefore, it is essential to consider correcting magnitude deviations before using these products. Based on independent assessment, this study suggests that further improvements are still needed for precipitation products on the QTP.
What problem does this paper attempt to address?